A graphical format is commonly used to display data sets representing various types of information. It may also be very useful to be able to view multiple data sets on a single graph. If the dependent variables of the data sets (generally represented on the vertical or y-axis of a graph) share a common range, all of the data sets may easily be displayed. Frequently, however, the ranges of the dependent variables are such that, if the data sets are displayed on one graph, some points may be lost because they are beyond the maximum or minimum values of the common y-axis. Alternatively, some information may not be useful to a viewer because of insufficient variability within a data set as it is displayed on the graph. That is, the maximum and minimum values of the common y-axis may greatly exceed the maximum and minimum values in the range of data points. FIG. 1 illustrates examples of the foregoing conditions. In the Fig., the first plot 102 represents a data set having maximum and minimum y-values which fit within the maximum and minimum values of the y-axis scale (0-30) and also have sufficient variability relative to the y-axis scale as to provide meaningful information to a viewer. The second plot 104 has maximum and minimum y-values which fit within the maximum and minimum values of the y-axis scale. However, there is insufficient variability relative to the y-axis scale to provide meaningful information to a viewer. Finally, the third plot 106 has maximum and minimum y-values which exceed the range of maximum and minimum values of the y-axis scale and therefore information is lost to the viewer.
One method of resolving the conflict between two data sets having significantly different ranges of y values is to provide two y-axes, one on each side of the graph and each having a range of values appropriate to one of the data sets. It will be appreciated, however, that such a graph will not accommodate more than two data sets having significantly different ranges of y values.
Another method of resolving a conflict is to separately scale the y-values of the data sets and display the plots against a y-axis. Scaling has generally been performed automatically by the computer on which the data plots are to be displayed. As such, the scaling may not provide a “best fit” for the data points for the viewer's purposes. FIG. 2 illustrates an example of three data sets 202, 204 and 206 plotted on a single graph against a common y-axis; the plot is a screen shot from the Microsoft® Windows® Performance viewer. The data sets are automatically scaled 208. However, two of the plots 202 and 206 exceed the common y-axis range and the third, while fully within the y-axis range, displays only minimal variability relative to the y-axis range and may provide little useful information to the viewer.
Consequently, a need remains for the ability to conveniently graphically display multiple data sets, having varying y-value ranges, on a common graph.